Autism is a devastating neuropsychiatric condition with unknown pathophysiology and for which there are no current effective treatments. Autism spectrum disorders are not uncommon and have an estimated incidence of approximately 1/1000. Although autism has a multifactorial etiology, it has a large genetic component. Recent genome scans have identified several potential chromosomal loci, but clear evidence for genetic heterogeneity has emerged, and the regions identified remain wide and in most cases only suggestive linkage to these regions has been found. Therefore, while a genetic approach to understanding autism etiology is likely to be fruitful, collaborative efforts involving large pooled samples will be necessary to achieve maximal power to identify disease critical regions narrow enough to permit positional cloning of autism susceptibility genes. A collaborative effort to produce an open data and biomaterials resource for the research community, the autism genetic resource exchange (AGRE), has been formed to facilitate the identification of autism susceptibility genes. This collaborative multi-site proposal seeks to expand and study the AGRE sample so as to identify and narrow autism susceptibility loci. Three hundred new multiplex families with autism spectrum disorders will be ascertained, for a total of 550 families in AGRE. A tiered, whole genome scan initially at 10 cM resolution, followed by fine mapping will be conducted to identify novel autism loci and more definitively confirm those previously identified. Phenotypic information will be used to help stratify the population so as to limit heterogeneity, as well as permit quantitative trait analysis focused on several heritable neurobehavioral components of autism. In parallel, karyotyping, and FISH using subtelomeric probes, and probes for several specific regions where chromosomal anomalies are highly associated with autism will be done. Finally, we have developed a novel, inexpensive, microarray-based method for SNP genotyping that will be used for association analyses, to permit narrowing of susceptibility regions identified. All phenotypic and genotype data will be made accessible via the Internet on a rolling basis, further enhancing the value of this resource to the community.